3. Impulse Sound Sources
111
summary in Popper 1980). A survey of the literature on odontocete acoustic
signals reveals a few early attempts to classify them (Kellogg et al. 1953;
Lilly and Miller 1961; Evans and Prescott 1962; Lilly 1962; Schevill 1964).
Generally, sounds were placed into two functional or three acoustic categories. The functional categories were echolocation and communication.
The acoustic categories were creaky buzzes (clicks), burst pulses, and
whistles. Most of the early attempts to classify signals acoustically were
based primarily upon how humans perceived these sounds. The functional
categories probably still apply, except that we might consider substituting
"social" in place of "communication," in light of the controversy that
normally accompanies the latter term.
In order to avoid some of these inherent pitfalls in our discussion of
sound generation mechanisms, we need explicit definitions of these sound
types. For the purposes of this chapter, we will distinguish sound types based
on the mechanism that probably produces them. Consequently, odontocete
sounds fall into three broad categories, (1) pulses or clicks, (2) whistles, and
(3) bangs. Pulses will be the primary focus of this chapter and subsumes two
of Lilly's (1962) categories, clicks (creaky buzzes) and burst pulses, since the
generation mechanism is probably the same, even though their functions
may be different. The reader should also recognize that all of these sound
categories could be further subdivided, but that is beyond the scope of this
chapter. Graphic representations of these basic sound types can be found
among the sonograms included in Tyack and Clark (Chapter 4).
Whistles are generally narrow bandwidth, often frequency modulated,
sounds that commonly last from half a second to a few seconds and may
have harmonic structure (Caldwell and Caldwell 1965; Norris et al. 1994).
Ridgway and his colleagues (Ridgway et al. 1980; Ridgway and Carder
1988) have identified nasal muscle groups that are active during whistle
production and have shown that air pressure in the nasal cavity increases
prior to the production of whistles. There is little else known about the exact
mechanism(s) of whistle generation because it has not often been the
primary subject of study. Generally, the frequency modulation, variable harmonic structure, and omnidirectional characteristics suggest that whistles
are primarily the result of vibrations that occur in air within the nasal
passages and diverticula of the odontocete head. (Amundin 1991b). These
"airborne" whistles may be coupled to tissues and/or water through
resonance phenomena (Lilly 1962). Mackay and Liaw (1981) proposed
that dolphin whistles might be produced by an "excited resonance" similar
to that in humans, but there is currently insufficient evidence to confirm
or refute this proposal. Ridgway and his colleagues (in preparation) have
confirmed that D. leucas can whistle at a depth of 300m, although there
are some associated changes in duration and frequency composition.
The function of odontocete whistles is thought to be primarily social
(Herman and Tavolga 1980), although there has been relatively less
study of them compared to pulsed sounds. There is some work that suggests
that unique whistles may function as individual "signatures" (Tyack and
111
summary in Popper 1980). A survey of the literature on odontocete acoustic
signals reveals a few early attempts to classify them (Kellogg et al. 1953;
Lilly and Miller 1961; Evans and Prescott 1962; Lilly 1962; Schevill 1964).
Generally, sounds were placed into two functional or three acoustic categories. The functional categories were echolocation and communication.
The acoustic categories were creaky buzzes (clicks), burst pulses, and
whistles. Most of the early attempts to classify signals acoustically were
based primarily upon how humans perceived these sounds. The functional
categories probably still apply, except that we might consider substituting
"social" in place of "communication," in light of the controversy that
normally accompanies the latter term.
In order to avoid some of these inherent pitfalls in our discussion of
sound generation mechanisms, we need explicit definitions of these sound
types. For the purposes of this chapter, we will distinguish sound types based
on the mechanism that probably produces them. Consequently, odontocete
sounds fall into three broad categories, (1) pulses or clicks, (2) whistles, and
(3) bangs. Pulses will be the primary focus of this chapter and subsumes two
of Lilly's (1962) categories, clicks (creaky buzzes) and burst pulses, since the
generation mechanism is probably the same, even though their functions
may be different. The reader should also recognize that all of these sound
categories could be further subdivided, but that is beyond the scope of this
chapter. Graphic representations of these basic sound types can be found
among the sonograms included in Tyack and Clark (Chapter 4).
Whistles are generally narrow bandwidth, often frequency modulated,
sounds that commonly last from half a second to a few seconds and may
have harmonic structure (Caldwell and Caldwell 1965; Norris et al. 1994).
Ridgway and his colleagues (Ridgway et al. 1980; Ridgway and Carder
1988) have identified nasal muscle groups that are active during whistle
production and have shown that air pressure in the nasal cavity increases
prior to the production of whistles. There is little else known about the exact
mechanism(s) of whistle generation because it has not often been the
primary subject of study. Generally, the frequency modulation, variable harmonic structure, and omnidirectional characteristics suggest that whistles
are primarily the result of vibrations that occur in air within the nasal
passages and diverticula of the odontocete head. (Amundin 1991b). These
"airborne" whistles may be coupled to tissues and/or water through
resonance phenomena (Lilly 1962). Mackay and Liaw (1981) proposed
that dolphin whistles might be produced by an "excited resonance" similar
to that in humans, but there is currently insufficient evidence to confirm
or refute this proposal. Ridgway and his colleagues (in preparation) have
confirmed that D. leucas can whistle at a depth of 300m, although there
are some associated changes in duration and frequency composition.
The function of odontocete whistles is thought to be primarily social
(Herman and Tavolga 1980), although there has been relatively less
study of them compared to pulsed sounds. There is some work that suggests
that unique whistles may function as individual "signatures" (Tyack and